Work

Individualized cerebellar damage predicts the presence of behavioral disorders in children with brainstem tumors

Year: 2025

Type: article

Abstract: Brainstem tumors often cause intractable neurobehavioral issues, which can be a challenge for patients and surgeons. Research on cerebellar changes in these patients is limited, despite symptoms similar to cerebellar injuries. This study aims to investigate cerebellar damage pattern resulting from brainstem tumors and its association with behavioral disorders. This study enrolled 147 children with brainstem tumors. A U-Net-based segmentation algorithm is used to divide their cerebellums into 26 lobules. And these lobules are then used to build a normative model for assessing individual structural deviations. Furthermore, a behavior prediction model is developed using the total outlier count (tOC) index and cerebellar lobule volume as predictive features. Over 95% of patients are found to have negative deviations in cerebellar regions, particularly in anterior lobules like Left V. Higher tOC is significantly associated with severe social problems (r = 0.31, p = 0.001) and withdrawal behavior (r = 0.28, p = 0.001). Smaller size of cerebellar regions strongly correlates with more pronounced social problems (r = 0.27, p = 0.007) and withdrawal behavior (r = 0.25, p = 0.015). Notably, lobules Right X, V, IV, VIIB, Left IX, VIII, and X influence social problems, while Left V, Right IV, Vermis VI, and VIII impact withdrawal behavior. Our study reveals cerebellar damage patterns in patients with brainstem tumors, emphasizing the role of both anterior and posterior cerebellar lobes in social problems and withdrawal behavior. This research sheds light on the cerebro-brainstem-cerebellar underlying complex behavioral disorders in brainstem tumor patients. Jia et al. employ a normative model to identify cerebellar damage patterns in children with brainstem tumors. Machine learning reveals that individualized cerebellar deviations predict social problems and withdrawn behaviours in these patients. The cerebellum is an important brain region that is associated with motor control and cognitive function. Brainstem tumors, which are located near the cerebellum, have been shown to impact cerebellum function. However, the extent of this impact and its association with behavioral disorders remain unclear. This study uses data from healthy children to develop a behavioral prediction model based on changes in the size of specific regions within the cerebellum using artificial intelligence. When applied to children with brainstem tumors, the results reveal that the anterior region of the cerebellum is most severely damaged. Additionally, overall cerebellar damage, especially in the posterior region of the cerebellum, is linked to social problems and withdrawal behaviors. This study highlights the model’s ability to assess brain structural changes and its potential to predict behavioral disorders based on the location of the cerebellum damage in patients with brainstem tumors.

Source: Communications Medicine

DOI: 10.1038/s43856-025-00810-0